We propose to apply technologies which have arisen in computer science and psychophysics to the improvement and evaluation of clinical performance as related to coronary arteries and coronary bypass grafts. Specifically, we intend to investigate how images from intravenous digital subtraction angiography and from digital subtraction root arteriography can be enhanced, and to compare the efficacy for lesion detection and quantitation of the enhanced images with unenhanced images and with those obtained by selective coronary arteriography. The proposed enhancement techniques include the use of standard image processing methods, and specifically developed techniques for vessel detection, three-dimensional blood-vessel reconstruction, modelling of myocardial dynamics, medial line representation, lumen size estimation, and display of vessel structure. We intend to implement our theoretical ideas into working software packages, which will then be evaluated by the techniques of Receiver Operating Characteristic (ROC) analysis applied to observer performance tests on images from a swine coronary atherosclerosis model.